def test_shap_rfe(X, y, capsys):
clf = DecisionTreeClassifier(max_depth=1)
with pytest.warns(None) as record:
shap_elimination = ShapRFECV(clf,
random_state=1,
step=1,
cv=2,
scoring='roc_auc',
n_jobs=4)
shap_elimination = shap_elimination.fit(X, y)
assert shap_elimination.fitted == True
shap_elimination._check_if_fitted()
report = shap_elimination.compute()
assert report.shape[0] == 3
assert shap_elimination.get_reduced_features_set(1) == ['col_3']
ax1 = shap_elimination.plot(show=False)
# Ensure that number of warnings was 0
assert len(record) == 0
# Check if there is any prints
out, _ = capsys.readouterr()
assert len(out) == 0
def test_shap_rfe_randomized_search(X, y, capsys):
clf = DecisionTreeClassifier(max_depth=1)
param_grid = {'criterion': ['gini'], 'min_samples_split': [1, 2]}
search = RandomizedSearchCV(clf, param_grid, cv=2, n_iter=2)
with pytest.warns(None) as record:
shap_elimination = ShapRFECV(search,
step=0.8,
cv=2,
scoring='roc_auc',
n_jobs=4,
verbose=150)
report = shap_elimination.fit_compute(X, y)
assert shap_elimination.fitted == True
shap_elimination._check_if_fitted()
assert report.shape[0] == 2
assert shap_elimination.get_reduced_features_set(1) == ['col_3']
ax1 = shap_elimination.plot(show=False)
# Ensure that number of warnings was at least 2 for the verbose (2 generated by probatus + possibly more by SHAP)
assert len(record) >= 2
# Check if there is any prints
out, _ = capsys.readouterr()
assert len(out) > 0
def test_shap_rfe_randomized_search_cols_to_keep(X, y, capsys):
"""
Test with ShapRFECV with column to keep param.
"""
clf = DecisionTreeClassifier(max_depth=1)
param_grid = {"criterion": ["gini"], "min_samples_split": [1, 2]}
search = RandomizedSearchCV(clf, param_grid, cv=2, n_iter=2)
with pytest.warns(None) as record:
shap_elimination = ShapRFECV(search,
step=0.8,
cv=2,
scoring="roc_auc",
n_jobs=4,
random_state=1)
report = shap_elimination.fit_compute(
X, y, columns_to_keep=["col_2", "col_3"])
assert shap_elimination.fitted
shap_elimination._check_if_fitted()
assert report.shape[0] == 2
reduced_feature_set = set(
shap_elimination.get_reduced_features_set(num_features=2))
assert reduced_feature_set == set(["col_2", "col_3"])
_ = shap_elimination.plot(show=False)
# Ensure that number of warnings was at least 2 for the verbose (2 generated by probatus + possibly more by SHAP)
assert len(record) >= 2
# Check if there is any prints
out, _ = capsys.readouterr()
assert len(out) == 0
def test_shap_rfe_cols_to_keep(X, y, capsys):
"""
Test for shap_rfe_cv with feautures to keep parameter.
"""
clf = DecisionTreeClassifier(max_depth=1, random_state=1)
with pytest.warns(None) as record:
shap_elimination = ShapRFECV(clf,
random_state=1,
step=2,
cv=2,
scoring="roc_auc",
n_jobs=4,
min_features_to_select=1)
shap_elimination = shap_elimination.fit(
X, y, columns_to_keep=["col_2", "col_3"])
assert shap_elimination.fitted
shap_elimination._check_if_fitted()
report = shap_elimination.compute()
assert report.shape[0] == 2
reduced_feature_set = set(
shap_elimination.get_reduced_features_set(num_features=2))
assert reduced_feature_set == set(["col_2", "col_3"])
# Ensure that number of warnings was 0
assert len(record) == 0
# Check if there is any prints
out, _ = capsys.readouterr()
assert len(out) == 0
def test_shap_rfe_svm(X, y, capsys):
"""
Test with ShapRFECV with SVM.
"""
clf = SVC(C=1, kernel="linear", probability=True)
with pytest.warns(None) as record:
shap_elimination = ShapRFECV(clf,
random_state=1,
step=1,
cv=2,
scoring="roc_auc",
n_jobs=4)
shap_elimination = shap_elimination.fit(X, y)
assert shap_elimination.fitted
shap_elimination._check_if_fitted()
report = shap_elimination.compute()
assert report.shape[0] == 3
assert shap_elimination.get_reduced_features_set(1) == ["col_3"]
_ = shap_elimination.plot(show=False)
# Ensure that number of warnings was 0
assert len(record) == 0
# Check if there is any prints
out, _ = capsys.readouterr()
assert len(out) == 0
def test_shap_rfe(X, y, sample_weight, capsys):
"""
Test with ShapRFECV.
"""
clf = DecisionTreeClassifier(max_depth=1, random_state=1)
with pytest.warns(None) as record:
shap_elimination = ShapRFECV(
clf,
random_state=1,
step=1,
cv=2,
scoring="roc_auc",
n_jobs=4,
)
shap_elimination = shap_elimination.fit(
X, y, sample_weight=sample_weight, approximate=True, check_additivity=False
)
assert shap_elimination.fitted
shap_elimination._check_if_fitted()
report = shap_elimination.compute()
assert report.shape[0] == 3
assert shap_elimination.get_reduced_features_set(1) == ["col_3"]
_ = shap_elimination.plot(show=False)
# Ensure that number of warnings was 0
assert len(record) == 0
# Check if there is any prints
out, _ = capsys.readouterr()
assert len(out) == 0